Shah Axat Bharatbhai

CM4050

Cement is widely used in the construction industry as it effectively binds building components like bricks, tiles, marbles, and stones to create concrete structures. However, the price of cement can be influenced by several factors, including the cost of raw materials, market demand, and supply capacity. Recent research reports indicate that cement prices in India may increase by around 3% due to Coal India's new coal pricing strategy. This rise in cement prices is also attributed to the high demand for cement in the market. The level of capacity utilization is another key factor that affects the cement price, with the Indian cement industry currently utilizing only around 80% of its capacity. To gain insight into the seasonal variations and trends in cement prices and to make accurate future predictions, researchers are using machine learning models. The literature review conducted provided valuable insights into the research area, which helped identify the research gap, formulate research questions and determine the research position. Previous studies have explored the use of classical regression models and modern AI systems to predict cement prices. To address this research gap, a Concurrent Nested Mixed research methodology was adopted, which involved collecting primary and secondary data. The primary data included daily cement prices, lead indicator data, and cement constituent prices, while the secondary data included data on the machine learning (ML) model used. Data analysis and validation were conducted using statistical tools to determine the correlation between variables, identify trends, analyze seasonal variations, evaluate the impact of holidays and outliers, and validate the accuracy of the FB Prophet model, which was utilized for the Time Series Analysis approach. The coding for the model was implemented using Python on Google Collab (Collaboratory), and the validation of the model was carried out to ensure the reliability and accuracy of the results. The primary objective of this research was to develop a reliable forecasting model for cement prices that can be utilized by contractors, dealers, manufacturers, and developers to gain insights into seasonal variations and facilitate effective resource planning and allocation. By developing suitable ML models based on the forecasted cement prices, stakeholders can benefit from accurate predictions and make informed decisions. The model's accuracy and reliability will enable the stakeholders to optimize their resource utilization and plan their activities efficiently.